SkillFoxx

GigaChat for Russian text editing: a Cursor skill and CLI

gigachat-ru-edit

A skill and CLI for Cursor: literary editing of Russian text through GigaChat without a rewrite and with a diff shown before applying

  • 0 ★ on the repository
  • Last commit
  • MIT
Skill

Extended access

Why this level

  • Texts are sent to an external Sber API, so confidential documents should not go this way
  • The GigaChat key sits in a file on disk and after a leak someone uses it in your name
All reasons and checks
Russian stack

gfermoto/gigachat-ru-editAuthor:Gfermoto

Install

Services

In your terminal, with SkillFoxx CLI

npx skillfoxx add skills/gigachat-ru-edit

Detects the agents on your machine, shows the access level and pins the version.

Other ways to install

Access and checks

How we review

Extended accessThe tool runs code or makes network calls and only reads external data.

  • Texts are sent to an external Sber API, so confidential documents should not go this way
  • The GigaChat key sits in a file on disk and after a leak someone uses it in your name

This is third-party code. Review the repository files before installing.

What it does

A Cursor skill with the `scripts/ru_edit.py` command wrapper edits Russian text through the GigaChat API. By default the `GigaChat-3-Ultra` model at `https://api.giga.chat/v1` is used with the free personal tier (scope `GIGACHAT_API_PERS`). The model fixes grammar and bureaucratese carefully without rewriting the text from scratch. The script takes one fragment through `--text` or a batch of JSON strings through `--json-file`. In `SKILL.md` the agent is told to collect Russian fragments, run the CLI, show a diff and apply edits only after the user agrees. Secrets live in `~/.config/gigachat.env`, and `GIGACHAT_PRESERVE_TERMS` protects terms that must not be touched.

Who it is for. Editors and developers who fix Russian texts and documentation in Cursor and want to use a Sber model for it.

Good fit when

  • Project docs and texts are full of bureaucratese and need a careful proofread
  • A batch of interface strings in JSON has to be proofread at once with a diff
  • Product names and terms must stay in Latin letters, untranslated
  • You want a free model for occasional edits with no foreign services

Not a fit when

  • The agent runs in Claude Code or Codex: the instructions and install path are written for the Cursor skills folder
  • You need a full rewrite of the style or a translation: the skill edits gently and does not write anew
  • The texts are confidential and must not go to an external cloud service

Example request

Proofread this documentation paragraph through GigaChat, remove bureaucratese and show a diff, and do not translate service names

Limitations

The skill is built for Cursor: installation is a clone into `~/.cursor/skills/gigachat-ru-edit`, and for other agents you port the commands yourself. A GigaChat authorization key from the Sber developer console and a Python environment with the dependencies from `requirements.txt` are required. Texts go to the Sber API, and the free personal tier limits are set by the service. Editing is done by a model, so the result should be reviewed through the diff rather than accepted blindly. The repository is tiny: seven files, zero stars, last commit in October 2026, MIT license.

How to disable. Delete the `~/.cursor/skills/gigachat-ru-edit` folder and the `~/.config/gigachat.env` file and revoke the key in the Sber developer console.

FAQ

Which model is used?

By default `GigaChat-3-Ultra` at `https://api.giga.chat/v1`, the Freemium tier for individuals.

Where are the keys stored?

In `~/.config/gigachat.env`: the `GIGACHAT_CREDENTIALS` variable with a `Basic ...` value and 600 file permissions.

Can many strings be edited at once?

Yes, through `--json-file`: a file with a list of objects like `{"id": "a", "text": "..."}`.

How do I protect terms from editing?

List them in the `GIGACHAT_PRESERVE_TERMS` variable and the model leaves them untranslated.

Are edits applied automatically?

No: per the instructions the agent shows a diff and applies changes only after confirmation.

Discussion

No questions or reviews yet.

All GigaChat tools
Russian stack

Eight skills for coding against the GigaChat API: choosing a layer, access setup, SDK, LangChain, gpt2giga and Deep Agents

SkillExtended access4
Russian stack

A skill collection for the Russian stack: GigaChat, YandexGPT, Yandex 360, Yandex Cloud, Metrika and marketplace orders

SkillExtended access82

OpenAI to GigaChat migration skill

migrate-to-gigachat-skill

Russian stack

A Cursor skill that migrates Python, JS or TS projects from OpenAI to GigaChat: detects the stack, picks an adapter, and verifies with tests it still starts

SkillExtended access1
Official

MCP server that checks text for signs of AI generation through the GigaChat API

MCP serverExtended access5
Official

An ai-forever proxy that serves GigaChat through OpenAI, Anthropic and Gemini APIs so Claude Code, Codex and Gemini CLI can run on it

CLIBroad access145
Russian stack

An MCP gateway to GigaChat and YandexGPT with automatic failover between them and 152-FZ personal-data anonymization

MCP serverExtended access1

README in short

The README sets out features, skill installation by cloning into the Cursor folder, getting a key and creating the environment file, running the CLI for one fragment and a JSON batch, and points to `SKILL.md` with agent instructions.

README badge

Report a problem